FACILITATORS

The content below contains a bracketed and highlighted placeholder for you to insert the information relevant to your institution. Don't forget to update it and remove this box before publishing the page.

It's important that we don't use learning data to assess academic integrity. Why? Because our data will never be 100% accurate (which, by the way, does not mean using those data isn’t incredibly powerful and important). We can’t get stuck in the trap that is thinking technology and data always accurately and objectively represent reality. They absolutely do not.

Emily Oakes, Principal Unizin IT Consultant, Indiana University

Addressing academic integrity concerns with data

There are numerous complexities impacting the effective use of activity logs and learning analytics to address academic integrity issues. The interpretation of data from learning tools can be influenced by the perspectives and biases of the individual(s) reviewing the data. Additionally, "the inability to track activity outside of an institutions' internal systems also impacts the ability to get a holistic picture of students' life-worlds." (Slade & Prinsloo, 2013). Simply put: even with the abundance of valuable data generated by learning tools, we don't have all of the data.

Further, learning management system data, like system data in general, can be both subject to capture issues (e.g. corruptions caused by access from an older device or unstable network connection) and human interpretation errors. Due to the potential inaccuracies of these logs, it is inappropriate to cite them as objective and accurate representations of behaviors when assessing academic integrity issues. More traditional methods of addressing these issues should be leveraged instead. In fact, Instructure's Canvas Guide about page views includes the following note:

Because mobile page view data is based on device settings and network connection, it may vary from the time the page views actually occurred. Page view data should not be used to assess academic integrity.

(Canvas Guide: How Do I View the Page Views for a User in an Account?)

For a real-life example of the potential consequences of using data to address academic integrity issues, the New York Times article Online Cheating Charges Upend Dartmouth Medical School describes a lawsuit that was brought against the university by students who were inappropriately accused of cheating based on Canvas activity data.

Using activity data to inform grading involves similar issues. Consider using traditional methods for assessing whether students are meeting course outcomes instead of using activity data or learning analytics as measures of engagement.

Diversity, Equity, Inclusion, & Justice tip

Often technology issues that arise will more greatly impact vulnerable students: those using screen readers, those who are only able to access older technology or mobile phones, etc. For example, for a few years Canvas didn't record page views when individuals were accessing the platform from a smartphone, causing Canvas to incorrectly report that students weren’t accessing their classes online. Gaps like these between web and mobile experiences disproportionately impact lower-income students:

As of early 2021, 27% of adults living in households earning less than $30,000 a year are smartphone-only internet users – meaning they own a smartphone but do not have broadband internet at home.

(Vogels, 2021)

Similarly, older devices unable to receive current software updates, unstable network connections (e.g. in rural areas), and more can impact the integrity of the data captured.

Get help

If you need assistance investigating and addressing a scholastic integrity issue, contact [insert institution contact information].